**From:** Biniam Tekle (*biniamt@dehai.org*)

**Date:** Thu Jan 29 2009 - 12:40:00 EST

Over the past few decades, seaport industry in many countries of the world

has witnessed remarkable

development. This is obvious, particularly in the East African countries,

such as Sudan, Eritrea,

Djibouti, Kenya, and Tanzania, and the Middle Eastern countries,

particularly Saudi Arabia, Yemen,

Oman, the United Arab Emirates, and Iran. These countries possess seaports

which are strategically

located in the international maritime trade route between the East and the

West (Figure 1) and are

considered as middle distance seaports. Goods carried from Europe and Far

East/Australia and vice

versa can be exchanged and transhipped to all countries in the Middle East,

Red *Sea*, and East Africa.

Since the olden days, these seaports have provided services for the regional

coasters and as time went

by, they have developed to be among the important maritime international

trade centres in the region.

The geographically strategic location of some of these seaports, have also

encouraged modern

container vessels to make short duration calls upon them (e.g. shipping

lines operating along

Asia/Europe route, Asia/Mediterranean route and Asia/US East Coast route).

These seaports and their

characteristics are displayed in Table 1.

------------------------------

*Page 3*

600

Ahmed Salem Al-Eraqi, Adli Mustafa, Ahamad Tajudin Khader and Carlos Pestana

Barros

*Table 1:*

Characteristics of seaports in Middle Eastern and East African regions

*No. Port *

*Berth Length *

*Equipment *

*Area M sq *

*Ship Call Total Tons *

1

Dubai Emirates (B)

5519

24

2209000

3916

12971235

2

Jeddah Saudi (B)

1330

26

50000

2049

12292704

3

Salalah Oman (B)

4296

14

341292

1506

1367404

4

Dammam Saudi (B)

1780

54

1032692

1653

19874564

5

Kuwait (B)

1750

23

538898

1636

3836840

6

Aden Yemen (B)

4875

176

1948610

6352

66541268

7

Mombasa Kenya (B)

4055

12

1586458

3148

16106155

8

Khor Fakkan Sharjah (B)

320

2

250567

398

1239645

9

Yanbu Saudi (M)

2004

34

843015

2463

14762086

10

Hodeidah (M)

1165

18

1321000

2042

8338290

11

Jubail Saudi (M)

8454

39

1843720

2782

16210109

12

Djibouti (M)

4800

9

1438800

1462

8556476

13

Dar es Salaam Tanzania (M)

1930

9

727000

1466

10720699

14

Sudan (M)

11200

114

2500000

4365

39245363

15

Mascut Oman (M)

2254

44

540253

2431

5102331

16

Asmara Eritrea (M)

1650

68

114117

1670

13916858

17

Khalid Sharjah (M)

2444

63

46864

1615

6232654

18

Bander Abbas Iran (M)

381

13

20000

195

334189

19

Mukalla Yemen (S)

385

6

400000

174

276681

20

Assab Eritrea (S)

1140

35

275319

819

535736

21

Tanga Tanzania (S)

1120

18

204057

1602

1509422

22

Mtwara Tanzania (S)

1795

20

151200

2165

6290892

B: Big port, M: Medium port and S: small port

*Figure 1: *Map of the region

------------------------------

*Page 4*

Efficiency of Middle Eastern and East African Seaports: Application of

DEA *Using* Window Analysis

601

*3. Literature Survey *

There is extensive literature on *DEA*, applied to a wide diversity of

economic fields and in particular to

seaports transportation. Cullinane et al. (2005) used *DEA *to highlight the

major objective of port

privatisation to improve the efficiency of this sector, with data of the

container throughput as output

and area and length terminal, quay crane, yard crane, straddle as inputs.

These authors concluded that

public and private/public *ports* perform better than public/private and

private *ports*.

Barros (2006) evaluated the performance of Italian seaports for the 2002

-2003 period *using*

*DEA *with Charnes, Cooper and Rhodes (*CCR*) model and Banker, Charnes and

Cooper (*BCC*) model,

to analyze 24 seaports. Barros (2006) used multiple efficiency models, such

as *DEA CCR*, *BCC*, Cross

efficiency *DEA *and *DEA *Super efficiency for Italian seaports, whereas

previously published articles

were limited to one or two analysis models. Because of this, the general

conclusion emerged that the

Italian companies display relatively high management skills, with most of

them being Variable Return

to Scale (*VRS*) efficient. Barros (2006) provides benchmarks to improve the

functioning of the port in

terms of efficiency.

Cullinane et al. (2004) applied window analysis in order to evaluate the

efficiency score of the

world's major container *ports* over time by *using* panel data and

cross-section data for 2003. They

concluded that the cross-section method is poor because it does not provide

details of port

performance, whereas the panel data with window analysis reflect a variation

of the absolute

performance of a port over time, and the relative performance of that port

in comparison to the others

at the same time.

Barros & Manolis (2004) compared the efficiency of *ports* of two European

countries, Greece

and Portugal. They took data from several *ports* of each of these countries

during the 1998-2000

periods. Their paper is intended to evaluate the efficiency of major

seaports in two small European

countries *using* the *CCR *and *BCC *models.

Wang & Cullinane (2006) focused on measuring the efficiency of container

terminals in

Europe. They proposed *DEA *with *CCR *and *BCC *models to evaluate

efficiency. They concluded that

management skills are crucial and emerge as a core in terms of business

competence.

Cullinane et al. (2006) contribute richly to this research by supporting

existing research which

leads to an estimates approach of relative efficiency in this active private

sector of *ports*. Their study

focuses on a sample comprising 69 of Europe's container terminals with

annual throughput of over

10,000 TEUs for the year 2002.

*4. Data Envelopment Analysis (DEA) *

*4.1. Standard DEA Models *

The basic concept of efficiency measurement is the ratio of total outputs to

total inputs. Charnes et al.

(1978) were the first to introduce the *DEA *as a multi-factor productivity

analysis module for measuring

the relative efficiencies on making units (*DMU*s). This model can not

support imperfectly competitive

markets. To overcome this limitation, Banker et al. (1984) described

*BCC *model,

this model estimates

its productivity level at the given scale of operation and identifies return

to scale. The goal is to select a

set of inputs and outputs that are relevant to the evaluation of performance

and for which a moderate

statistical relationship exists.

In *DEA*-*CCR *model all observed production combinations can be scaled up

or down

proportionally, and in *DEA*-*BCC *model the variables allow return to scale

and is graphically

represented by a piecewise linear convex frontier (Cullinane et al. 2006).

The *DEA *is normally applied

to analyse the cross section data, where time is ignored and *DMU *are

compared with the others at the

same period. In this paper, we propose the output-oriented *DEA *model to

maximize the output while

the given current inputs remain the same. The mathematical expression of the

*DEA *models as follow:

1) *CCR *Model (Charnes, Cooper and Rhodes) (1978).

------------------------------

*Page 5*

602

Ahmed Salem Al-Eraqi, Adli Mustafa, Ahamad Tajudin Khader and Carlos Pestana

Barros

*Max *

*k*

φ

*ik*

*ij*

*n*

*j*

*j*

*x*

*x*

*ts*

≥

∑

=1

..

λ

*i=1, 2… m;*

(1)

*rk*

*k*

*rj*

*n*

*j*

*j*

*y*

*y*

φ

λ

≤

∑

=1

*r=1,2,…,s;*

0

≥

*j*

λ

*j*

∀

*.*

And 2) *BCC *Model, (Banker, Charnes and Cooper 1984) is defined by adding

equations (2) to

expression (1) above.

1

1

=

∑

=

*n*

*j*

*j*

λ

(2)

Where *n *is number of *DMU*,

*k*

φ

is the efficiency of the kth *DMU*, *x*

*ij*

are *i-th *inputs of the *j-th*

*DMU*, *y*

*rj*

are the outputs of *j-th DMU *and *λ*

*j*

is weight of *j-th DMU*. The *DEA*-technique requires a

large number of medium-sized linear programming problems to be solved. The

two models, described

previously, the first is called *CCR *model (constant return to scale) which

is a scale efficiency and

technical efficiency, and the second is called *BCC *model (variable return

to scale) which is a pure

technical and scale efficiency (Fare et al. 1994). That output-oriented

efficiency problem can be written

in the form of N linear programming system (Cullinane et al. 2004). The

technical efficiencies derived

from the *DEA*-*CCR *and *DEA*-*BCC *models are frequently used to obtain a

measure of scale for *DMU*,

given by SE

k

=U

*CCR*k

/ U

*BCC*k

(William et al.2000), where U

*CCR*_k

and U

*BCC*_k

are the technical efficiency

measures for *DMU k *derived from applying the *DEA*-*CCR *and

*DEA*-*BCC *models

respectively. *CCR*

score is called technical efficiency (TE), *BCC *called pure technical

efficiency (PTE), and scale

efficiency noted by (SE) with TE = PTE * *SE*, if *SE*

*k*

*=*1 then the score is efficiency (constant return to

scale) otherwise the score is inefficiency if SE

k

<1(Increasing or decreasing return to scale). The

constant return to scale means that the firm able to operate the inputs and

outputs linearly without

increasing or decreasing. The increasing return to scale means that the firm

operating at lower scale

sizes, while decreasing return to scale means that the firm operating at

higher scale sizes.

*4.2. Window Analysis *

A *DEA *window analysis calculates the average efficiency of *CCR *and *BCC

*models, and is useful for

detecting efficiency trends of unit over time (Charnes et al., 1994b). In

such a circumstance, *DEA*

window analysis can be adopted to detect a trend of *DMU *over time (Asmild

et al. 2004; Charnes et al.

1994b; Yue, 1992). The procedure is to consider each *DMU *is represented as

if it were different *DMU*

in each period under analysis. There is no theory underpins the

justification for the choice of window

size. The common notation is describing as follow:

*n*=Number of *DMU*

*T*=Number of periods

*p*= Altitude of window

(*p*

≤

*T*) (Number of columns in window)

*w*= Number of windows

(Number of rows for each firm)

where *w=T-p+1 *are number of analysis for each *DMU *and *n x p *will be

the total number analysis for

all *DMUs *as mention above. The identification of performance trends in row

window and the stability

is defined in column. The variation in row reflects both the absolute

performance of a port over time

and the relative performance of that port in comparison to the others *ports

*.

------------------------------

*Page 6*

Efficiency of Middle Eastern and East African Seaports: Application of

DEA *Using* Window Analysis

603

*5. Data and Statistical Analysis *

*5.1. The input and output measures *

The data were obtained from the annual statistics reports of

*ports*authorities, by fax and E-mail and

through internet (*using* Google Earth and *ports* web site as

Maritimechain.com and *Ports* Harbours

Marines Worldwide). The inputs remain unchanged within this time; by

contrast, change occurred in

the cargo throughput and ship calls.

To estimate the efficiency of the *ports* under study, we used data for the

years 2000-2005; the

*ports* considered in analysis are listed in Table 1 and the summary of

their characteristics are described

in Table 2.

The output is measured by two indicators: 1) Ship calls, and 2) Throughput

(movement of

general cargo dry and liquids and containers) load/unload, while the inputs

are measured by the

indicators, such as berth length, *storage* area, and handling equipment.

*Table 2:*

Summary statistics for years 2000-2005

*Inputs *

*Outputs *

*Berth *

*Length(m) *

*Storage *

*Area(m*

*2*

*) *

*Handling*

*Equipment *

*Ship Calls *

*(Units) *

*Throughput (Tons) *

Mean

2938.500

37.318

835584.636

2086.606

12102800.015

Std. Error of Mean

232.772

3.446

66339.004

125.965

1539146.311

Median

1862.500

23.500

539575.500

1818.500

6831638.500

Mode

320.000

9.000

20000.000

1450.000

241950.000

Std. Deviation

2674.343

39.590

762177.126

1447.233

17683444.807

Variance

7152111.641

1567.364

580913971516.997

2094483.080

312704220232032.000

Skewness

1.664

2.233

0.764

1.308

3.929

Kurtosis

0.211

0.211

0.211

0.211

0.211

Range

2.404

4.956

-0.726

2.290

20.790

Minimum

0.419

0.419

0.419

0.419

0.419

Maximum

10880.000

174.000

2480000.000

7450.000

129429309.000

Sum

320.000

2.000

20000.000

124.000

63644.000

Count

22

22

22

22

22

*5.2. Correlation and regression analysis *

The analyses of inputs and outputs variables data show only those that are

highly interrelated (refer to

Table 3).

*Table 3:*

Correlation coefficients with inputs and outputs

*Berth Length Handling Equipment*

*Storage Area*

*Ship Calls *

*Throughput *

Berth Length

1.000*

Handling Equipment

0.469

1.000*

*Storage* Area

0.812*

0.434

1.000*

Ship Calls

0.664*

0.762*

0.679*

1.000*

Throughput

0.561*

0.896*

0.663*

0.879*

1.000*

*Correlation coefficient values are significant at the 0.05 level

(2-tailed).

The multiple regressions are used to determine any relationship between the

inputs and the

outputs. Table 4 shows the "R2" values as the proportion of variation in the

dependent variable ship

calls and throughputs explained by the regression model are 0.801 and 0.907.

The statistics and its

significant values are used to test the null hypothesis that the regression

coefficient is zero that mean

there is a linear relationship between the dependent (ship calls and

throughput) and independent (berth

length, equipment and area) variables.

------------------------------

*Page 7*

604

Ahmed Salem Al-Eraqi, Adli Mustafa, Ahamad Tajudin Khader and Carlos Pestana

Barros

*Table 4:*

Regression results on inputs and output variables

*Outputs *

*Inputs *

*Ship Calls *

*Throughput *

Berth Length

-0.015

-1082.928

Handling Equipment

19.015

290517.165

*Storage* Area

0.001

9.391

Constant

592.118

-3403512.341

R

2

0.801

0.907

The software Efficiency Measurement System version 1.3 from Holger Scheel

was applied to

solve the *DEA *with two models on the return to scale of *ports* production

function, called *CCR *model

(constant return to scale) and *BCC *model (variable return to scale).

*6. Results *

We first applied *DEA *to analyse the efficiency score of the *ports*, we

computed efficiency *using* two

models: *DEA*-*CCR *and *DEA*-*BCC*. *DEA *is carried on 22 *ports* shown in

Table 1. Table 5 represents the

efficiency estimates, the scale efficiency and scale type of each port. The

score report shows that 7 and

9 *ports* out 22 are efficient under *DEA*-*CCR *and *DEA*-*BCC *models,

respectively. The results of two

models show that the number of efficient *ports* in *BCC *is more than *CCR

*with average values of 0.786

and 0.875, respectively.

The output oriented model was applied in this paper to select the *ports* in

terms of berth length,

*storage* area and handling equipment. Theatrically, the output of technical

efficiency is given by

*TE*

k

=1/U

k

for k term of *DMU *(U

k

is an inefficient score under *CCR **using* output-oriented). The *ports*

under study must increase their product on an average of 1.272 times for the

same inputs. The scale

properties of *ports* production show 7 *ports* constant return to scale, 8

increasing return to scale, 7

decreasing return to scale. Note that 7 *ports*, Khor Fakkan, Dubai, Kuwait,

Mukalla, Hodeidah, Yanbu,

and Djibouti are efficient under *CCR *and *BCC*.

We next applied *DEA *window to analyse the efficiency score of the *ports*,

with two models

*DEA*-*CCR *and *DEA*-*BCC*. The window analysis is used to examine the

efficiency over time for the

period 2000-2005 (6 years x 22 *ports* = 110 observations), T=6, p=3 and

w=4. *DEA *is carried on 22

*ports* shown in Table 1. As such, the length of the window used here is

defined as three (Charnes et al.

1985). the scale efficiency of each port. Four separate windows are

represented as separate rows in

Tables 6 and 7. Tables 6 and 7 represent the efficiency estimates, the

average of DEA efficiency scores

and its standard deviation in the columns denoted 'Mean' and 'S.D'.

The identification of performance trends in row window and the stability is

defined in column

of each year that allows controlling both of them through the separate

windows. The efficiency score

estimated shows that 16 and 17 *ports* are stable (have low standard

deviation) under *CCR *and *BCC*,

respectively, on the other hand, 6 and 5 *ports* are unstable (have high

standard deviation) under *CCR*

and *BCC*, respectively. The efficiency score mean value shows better under

*BCC *than *CCR*, although

all the *ports* still inefficient and reflect a fluctuation in efficiency

score. There is an improvement in the

efficiency for Khor Fakkan, Kuwait, and Djibouti *ports* with *CCR*, and

Bander Abbas, Khor Fakkan,

Dubai, Kuwait, Mukalla, Mombasa and Djibouti *ports* with *BCC*. The

variation haphazard

(increasing/decreasing or decreasing/increasing) in performance impacted the

main efficiency over the

time period. In general, all the *ports* are stable. Table 8 shows the scale

efficiency of all the *ports* over

the entire time periods of study.

The comparison between cross-section data and panel data in Tables 5 and in

Tables 6, 7 shows

a similarity in average efficiency score for most of these *ports*.

------------------------------

*Page 8*

Efficiency of Middle Eastern and East African Seaports: Application of

DEA *Using* Window Analysis

605

*7. Discussion *

In this paper *DEA *cross-section data and window analysis are used to

determine the relative efficiency

of 22 cargo *ports* in the Middle East and Africans countries. The results

of cross-section provide

information for the overall time period. The panel data provide large

details of performance analysis

over a period of time. The fluctuation of the efficiency score with window

analysis, due to the

comparison between the big *ports* which have high production and small *

ports* which have low

production. This study shows that small *ports* are efficient while big *

ports* are inefficient. The

indicators of production scale in this study as shown in Table 6 and 7 are

the main factors of efficiency

and inefficiency.

The inefficiency of *ports* may also have resulted for reasons, such as 3

rd

Gulf War and to other

reasons related to the security of ship companies particularly in this

region during 2003-2004. We

conclude that for increasing port efficiency, ships arrival should be

encouraged to increase the scale of

production; on the other hand, the inefficient *ports* with declining

efficiency reduce their scale of

operation to be efficient. The comparison of the two methods shows biases in

efficiency over the time

for Dubai, Mukalla, Hodeidah and Yanbu under CCR and Hodeidah, Yanbu under

BBC, respectively.

This result provided with window analysis discusses the recent changes of

performance and

stability of the port over time.

*Table 5:*

The relative efficiency of seaports *using* *DEA*-*CCR *and *DEA*- *BCC *models

in 2001

*Country Port *

*DEA - CCR *

*DEA - BCC *

*Scale Efficiency *

*Return to scale *

Bander Abbas Iran

0.803

1.000

0.803

Decreasing

Khor Fakkan Sharjah

1.000

1.000

1.000

Constant

Khalid Sharjah

0.834

0.848

0.983

Increasing

Salalah Oman

0.918

0.934

0.983

Increasing

Mascut Oman

0.683

0.726

0.941

Decreasing

Dubai Emirates

1.000

1.000

1.000

Constant

Kuwait

1.000

1.000

1.000

Constant

Mukalla Yemen

1.000

1.000

1.000

Constant

Aden Yemen

0.862

0.953

0.904

Decreasing

Hodeidah

1.000

1.000

1.000

Constant

Dammam Saudi

0.515

0.725

0.711

Decreasing

Jubail Saudi

0.708

0.735

0.964

Increasing

Yanbu Saudi

1.000

1.000

1.000

Constant

Jeddah Saudi

0.526

0.840

0.626

Decreasing

Sudan

0.683

0.862

0.792

Decreasing

Mombassa Kenya

0.876

0.985

0.889

Decreasing

Dar es Salaam Tanzania

0.841

0.870

0.966

Increasing

Tanga Tanzania

0.331

1.000

0.331

Increasing

Mtwara Tanzania

0.261

0.331

0.789

Increasing

Assab Eritrea

0.454

0.460

0.987

Increasing

Asmara Eritrea

0.989

0.989

1.000

Increasing

Djibouti

1.000

1.000

1.000

Constant

Average

0.786

0.875

0.894

------------------------------

*Page 9*

606

Ahmed Salem Al-Eraqi, Adli Mustafa, Ahamad Tajudin Khader and Carlos Pestana

Barros

*Table 6:*

DEA-CCR window analysis for cargo port efficiency (100='efficient')

*Efficiency Scores *

*Summary Measures *

*Port*

*2000*

*2001*

*2002*

*2003*

*2004*

*2005*

*Mean *

*S. D.*

Bander

67.101

73.811

68.539

Abbas

71.300

66.207

64.423

66.882

65.080

66.345

62.831

64.052

68.768

67.111

3.129

Khor

90.640

97.756

100.000

Fakkan

97.756

100.000

97.516

Sharjah

100.000

97.516

100.000

93.498

92.995

100.000

97.306

3.224

Khalid

69.749

88.855

95.933

Sharjah

76.323

82.403

59.862

80.665

58.599

61.503

53.922

56.594

94.932

73.278

15.289

Salalah

26.373

29.326

44.516

Oman

29.326

44.516

96.202

51.151

90.736

60.580

84.196

63.741

99.573

60.020

26.954

Mascut

44.747

51.780

65.175

Oman

47.543

59.841

60.494

61.166

62.056

65.836

58.799

62.381

62.285

58.508

6.794

Dubai

53.026

63.466

67.141

63.466

67.141

73.917

74.111

82.503

100.000

84.964

100.000

93.775

76.959

15.334

Kuwait

83.692

88.705

100.000

72.271

83.293

100.000

83.293

100.000

93.758

100.000

96.596

100.000

91.801

9.364

Mukalla

100.000

67.376

63.905

Yemen

71.588

71.588

74.913

76.587

80.566

82.385

76.780

77.920

100.000

78.634

11.265

Aden

60.713

82.786

74.605

Yemen

81.824

73.739

51.281

84.716

65.301

66.361

63.829

65.036

73.488

70.306

10.003

Hodeidah

53.286

100.000

89.660

Yemen

100.000

89.660

45.157

100.000

52.105

52.449

66.915

66.945

71.653

73.986

20.952

Dammam

31.622

46.264

48.696

42.238

44.459

33.832

43.981

33.746

47.459

48.492

47.789

60.458

44.086

8.031

------------------------------

*Page 10*

Efficiency of Middle Eastern and East African Seaports: Application of

DEA *Using* Window Analysis

607

*Table 6:*

Continued

*Efficiency Scores *

*Summary measures *

*Port*

*2000*

*2001*

*2002*

*2003*

*2004*

*2005*

*Mean *

*S. D.*

Jubail

46.310

95.000

100.000

Saudi

95.000

100.000

39.770

100.000

39.770

55.600

39.770

55.600

76.246

70.255

26.467

Yanbu

80.026

94.000

100.000

94.000

100.000

47.190

100.000

47.448

56.096

45.630

85.462

62.511

76.030

22.618

Jeddah

29.590

29.714

32.182

29.136

31.075

39.391

31.075

39.121

38.889

45.708

45.489

66.746

38.176

10.843

Sudan

69.583

46.165

71.068

46.165

71.068

52.799

73.213

54.394

74.386

51.850

70.232

48.949

60.823

11.558

Mombasa

65.910

62.270

52.749

62.270

52.749

88.780

52.749

88.780

58.148

85.122

55.751

47.767

64.420

14.855

Dar es

67.154

89.369

73.981

Salaam

89.369

73.981

87.177

73.981

87.177

83.560

83.585

80.117

88.519

81.497

7.529

Tanga

30.364

30.529

28.714

30.529

28.714

35.150

28.784

35.235

25.310

33.553

24.101

40.641

30.969

4.581

Mtwara

20.946

18.390

12.194

18.359

12.174

13.941

14.449

16.527

27.565

20.200

33.116

20.585

19.037

6.227

Assab

33.147

35.249

38.134

35.249

38.134

44.538

39.294

45.894

40.806

42.600

37.877

48.570

39.958

4.673

Asmara

64.047

100.000

82.910

96.442

79.910

88.231

80.856

89.275

88.606

76.499

76.054

100.000

85.236

10.706

Djibouti

90.000

98.995

100.000

79.212

80.017

100.000

79.092

100.000

100.000

91.688

89.609

100.000

92.384

8.794

------------------------------

*Page 11*

608

Ahmed Salem Al-Eraqi, Adli Mustafa, Ahamad Tajudin Khader and Carlos Pestana

Barros

*Table 7:*

*DEA-BCC *window analysis for cargo port efficiency (100='efficient')

*Efficiency Scores *

*Summary Measures *

*Port*

*2000*

*2001*

*2002*

*2003*

*2004*

*2005*

*Mean *

*S. D.*

Bander

90.963

100.000

94.027

Abbas

100.000

93.489

91.288

Iran

95.482

92.910

94.715

91.367

93.205

100.000

94.787

3.423

Khor

90.640

97.756

100.000

Fakkan

97.756

100.000

97.516

Sharjah

100.000

97.516

100.000

93.498

92.995

100.000

97.306

3.224

Khalid

70.151

89.367

96.486

Sharjah

78.884

85.168

61.871

83.436

60.612

63.615

54.526

57.228

95.995

74.778

15.272

Salalah

31.556

35.089

50.860

Oman

32.295

49.408

100.000

51.281

94.244

62.024

84.566

64.217

100.000

62.962

25.870

Mascut

47.347

54.790

68.962

Oman

49.878

62.780

63.465

64.277

65.041

68.962

64.501

68.384

68.311

62.225

7.471

Dubai

78.977

94.527

100.000

Emirates

85.861

90.833

100.000

84.909

93.478

100.000

93.478

100.000

93.775

92.987

6.857

Kuwait

85.051

89.360

100.000

72.271

84.080

100.000

84.614

100.000

100.000

100.000

96.596

100.000

92.664

9.350

Mukalla

100.000

100.000

80.610

Yemen

100.000

95.562

100.000

93.352

99.591

100.000

98.563

98.969

100.000

97.221

5.644

Aden

73.885

100.000

90.982

Yemen

95.302

85.885

59.728

91.801

67.378

68.312

68.946

69.940

80.496

79.388

13.115

Hodeidah

53.286

100.000

89.660

Yemen

100.000

89.660

45.157

100.000

52.258

52.632

67.923

67.923

72.701

74.267

20.853

Dammam

50.800

72.533

76.347

Saudi

68.040

71.618

55.430

71.105

55.326

76.228

50.416

75.837

73.053

66.394

10.270

------------------------------

*Page 12*

Efficiency of Middle Eastern and East African Seaports: Application of

DEA *Using* Window Analysis

609

*Table 7: *Continued

*Efficiency Scores *

*Summary Measures *

*Port*

*2000*

*2001*

*2002*

*2003*

*2004*

*2005*

*Mean *

*S. D.*

Jubail

48.728

95.000

100.000

Saudi

95.000

100.000

41.032

100.000

40.996

57.314

40.229

56.242

77.127

70.972

25.865

Yanbu

80.026

94.000

100.000

Saudi

94.000

100.000

47.986

100.000

47.986

56.191

46.029

93.637

63.057

76.909

22.806

Jeddah

75.111

79.286

83.276

Saudi

70.984

76.238

97.236

70.329

72.351

75.872

72.024

75.529

80.760

77.416

7.390

Sudan

93.729

62.186

95.729

57.247

88.127

65.473

87.313

64.869

88.711

61.612

84.258

57.424

75.556

15.182

Mombasa

100.000

100.000

94.916

Kenya

85.214

86.237

100.000

80.427

100.000

95.947

100.000

88.092

79.873

92.559

8.067

Dar es

69.572

92.587

76.645

Salaam

92.444

76.527

90.177

Tanzania

76.527

90.177

86.436

86.298

82.717

91.392

84.291

7.756

Tanga

99.460

100.000

100.000

Tanzania

86.855

86.978

100.000

86.978

100.000

71.831

82.558

59.302

100.000

89.497

13.289

Mtwara

28.715

25.210

16.717

Tanzania

33.453

22.182

25.403

22.927

26.255

42.711

21.898

35.623

22.361

26.955

7.168

Assab

33.753

35.894

38.831

Eritrea

36.430

39.411

46.030

40.470

47.267

42.027

42.767

38.025

48.784

40.807

4.717

Asmara

64.054

100.000

82.918

Eritrea

100.000

82.893

91.512

84.008

92.757

92.062

76.540

76.112

100.000

86.905

11.213

Djibouti

90.000

98.995

100.000

79.212

80.017

100.000

79.092

100.000

100.000

92.045

89.640

100.000

92.417

8.791

------------------------------

*Page 13*

610

Ahmed Salem Al-Eraqi, Adli Mustafa, Ahamad Tajudin Khader and Carlos Pestana

Barros

*Table 8:*

Score Efficiency (1='efficient')

*Scores Efficiency*

*(return to scale)*

*Port *

*2000 *

*2001 *

*2002 *

*2003 *

*2004 *

*2005 *

Bander Abbas 0.74 (Decreasing) 0.74 (Decreasing)

0.73 (Decreasing)

0.71 (Decreasing) 0.71 (Decreasing) 0.71 (Decreasing)

0.70 (Decreasing) 0.70 (Decreasing) 0.70 (Decreasing)

0.69 (Decreasing) 0.69 (Decreasing) 0.69 (Decreasing)

Khor

1.00(Constant)

1.00(Constant)

1.00(Constant)

Fakkan

1.00(Constant)

1.00(Constant)

1.00(Constant)

Sharjah

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

Khalid

0.99(Inc)

0.99(Increasing)

0.99(Increasing)

Sharjah

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.99(Increasing)

0.99(Increasing)

0.99(Increasing)

Salalah

0.84(Increasing)

0.84(Increasing)

0.88(Increasing)

Oman

0.91(Increasing)

0.90(Increasing)

0.96(Increasing)

1.00(Constant)

0.96(Increasing)

0.98(Increasing)

1.00(Constant)

0.99(Increasing)

1.00(Constant)

Mascut

0.95(Increasing)

0.95(Increasing)

0.95(Increasing)

Oman

0.95(Increasing)

0.95(Increasing)

0.95(Increasing)

0.95(Increasing)

0.95(Increasing)

0.95(Increasing)

0.91(Increasing)

0.91(Increasing)

0.91(Increasing)

Dubai

0.67(Decreasing)

0.67(Decreasing)

0.67(Decreasing)

0.74(Decreasing)

0.74(Decreasing)

0.74(Decreasing)

0.87(Increasing)

0.88(Decreasing)

1.00(Constant)

0.91(Decreasing)

1.00(Constant)

1.00(Constant)

Kuwait

0.98(Increasing)

0.99(Increasing)

1.00(Constant)

1.00(Constant)

0.99(Increasing)

1.00(Constant)

0.98(Increasing)

1.00(Constant)

0.94(Decreasing)

1.00(Constant)

1.00(Constant)

1.00(Constant)

Mukalla

1.00(Constant)

0.67(Increasing)

0.79(Increasing)

Yemen

0.72(Increasing)

0.75(Increasing)

0.75(Increasing)

0.82(Increasing)

0.81(Increasing)

0.82(Increasing)

0.78(Increasing)

0.79(Increasing)

1.00(Constant)

Aden

0.82(Increasing)

0.83(Decreasing)

0.82(Decreasing)

0.86(Decreasing)

0.86(Decreasing)

0.86(Increasing)

0.92(Decreasing)

0.97(Increasing)

0.97(Increasing)

0.93(Increasing)

0.93(Increasing)

0.91(Increasing)

Hodeidah

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

0.99(Increasing)

0.99(Increasing)

0.99(Increasing)

Dammam

0.62(Increasing)

0.64(Decreasing)

0.64(Decreasing)

0.62(Decreasing)

0.62(Decreasing)

0.61(Increasing)

0.62(Decreasing)

0.61(Increasing) 0.62(Decreasing)

0.96(Increasing) 0.63(Decreasing)

0.83(Increasing)

------------------------------

*Page 14*

Efficiency of Middle Eastern and East African Seaports: Application of

DEA *Using* Window Analysis

611

*Table 8: *Continued

*Port *

*2000 *

*2001 *

*2002 *

*2003 *

*2004 *

*2005 *

Jubail

0.95

1.00(Constant)

1.00(Constant)

Saudi

1.00(Constant)

1.00(Constant)

0.97(Increasing)

1.00(Constant)

0.97(Decreasing)

0.97(Increasing)

0.99(Increasing)

0.99(Increasing)

0.99(Increasing)

Yanbu

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

0.98(Increasing)

1.00(Constant)

0.99(Increasing)

1.00(Constant)

0.99(Increasing)

0.91(Decreasing)

0.99(Increasing)

Jeddah

0.39(Dec)

0.37(Dec)

0.39(Dec)

0.41(Decreasing)

0.41(Decreasing)

0.41(Decreasing)

0.44(Decreasing)

0.54(Decreasing)

0.51(Decreasing)

0.63(Decreasing)

0.60(Decreasing)

0.83(Increasing)

Sudan

0.74(Decreasing)

0.74(Increasing)

0.74(Decreasing)

0.81(Increasing)

0.81(Decreasing)

0.81(Increasing)

0.84(Decreasing)

0.84(Increasing)

0.84(Decreasing)

0.84(Increasing)

0.83(Decreasing)

0.85(Increasing)

Mombasa 0.66(Decreasing)

0.62(Decreasing)

0.56(Decreasing)

0.73(Decreasing)

0.61(Decreasing)

0.89(Decreasing)

0.66(Decreasing)

0.89(Decreasing)

0.61(Decreasing)

0.85(Decreasing)

0.63(Decreasing)

0.60(Decreasing)

Dar es

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

Salaam

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

Tanga

0.31(Decreasing)

0.31(Decreasing)

0.29(Decreasing)

0.35(Decreasing)

0.33(Decreasing)

0.35(Decreasing)

0.33(Decreasing)

0.35(Decreasing)

0.35(Decreasing)

0.41(Decreasing)

0.41(Decreasing)

0.41(Decreasing)

Mtwara

0.73(Increasing)

0.73(Increasing)

0.73(Increasing)

0.55(Increasing)

0.55(Increasing)

0.55(Increasing)

0.63(Increasing)

0.63(Increasing)

0.65(Increasing)

0.92(Increasing)

0.93(Increasing)

0.92(Increasing)

Assab

0.98(Increasing)

0.98(Increasing)

0.98(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

0.97(Increasing)

1.00(Constant)

1.00(Constant)

1.00(Constant)

Asmara

1.00(Constant)

1.00(Constant)

1.00(Constant)

0.96(Increasing)

0.96(Increasing)

0.96(Increasing)

0.96(Increasing)

0.96(Increasing)

0.96(Increasing)

1.00(Constant)

1.00(Constant)

1.00(Constant)

Djibouti

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

1.00(Constant)

*Acknowledgment *

The authors are grateful to the *ports* authorities for providing data and

information.

------------------------------

*Page 15*

612

Ahmed Salem Al-Eraqi, Adli Mustafa, Ahamad Tajudin Khader and Carlos Pestana

Barros

*References *

[1]

Asmild, M., J.C. Paradi, V. Aggarwall and C. Schaffnit, 2004. "Combining DEA

Window

Analysis with the Malmquist Index Approach in a Study of the Canadian

Banking Industry",

*Journal of Productivity Analysis *21, pp. 67–89.

[2]

Banker R. D., Charnes A. and Cooper W. W. (1984). Model for Estimating

Technical and Scale

Efficiencies in Data Envelopment Analysis. Management Science 30(9):pp.

1078-1092.

[3]

Barros, C.P., 2006. "A Benchmark Analysis of Italian Seaports *using* Data

Envelopment

Analysis", Maritime *Economics & Logistics *8, pp. 347–365.

[4]

Barros, C.P. and A. Manolis, 2004. "Efficiency in European Seaports with

DEA: Evidence

from Greece and Portugal", *Maritime Economics & Logistics *6, pp. 122–140.

[5]

Charnes, A., W.W. Cooper and E. Rhodes, 1978. "Measuring the Efficiency of

Decision

Making Units" *European Journal of Operational Research *2(6), pp. 429–444.

[6]

Charnes, A., C.T. Clark, W.W. Cooper, and B. Golany, 1985. "A Developmental

Study of Data

Envelopment Analysis in Measuring the Efficiency of Maintenance Units in the

US Air Force"

*Annals of Operations Research *2, pp. 95–112.

[7]

Charnes, A., W.W. Cooper, A.Y. Lewin and L.M. Seiford, 1994a. *Data

Envelopment Analysis: *

*Theory, Methodology and Application. *Norwell: Kluwer Academic Publishers.

[8]

Charnes, A., W.W. Cooper and L.M., 1994b. "Extension to DEA Models" In A.

Charnes, W.

W. Cooper, A. Y. Lewin, & L. M. Seiford (Eds.). *Data Envelopment Analysis:

Theory,*

*Methodology and Applications. *Norwell: Kluwer Academic Publishers.

[9]

Cullinane, K., D.W. Song and R. Gray, 2002. "A Stochastic Frontier Model of

the Efficiency of

Major Container Terminals in Asia: Assessing the Influence of Administrative

and Ownership

Structures" *Transportation Research Part A *36, pp. 734–762.

[10]

Cullinane, K., P. Ji and T. Wang, 2004. "An Application of DEA Windows

Analysis to

Container Port Production" *Review of Network Economics *3(2), pp. 184–206.

[11]

Cullinane, K., D. Song and T. Wang, 2005. "The Application of Mathematical

Programming

Approaches to Estimating Container Port Production Efficiency" *Journal of

Productivity *

*Analysis *24, pp. 73–92.

[12]

Cullinane K., P. Ji and T. Wang, 2006. "The Efficiency of European Container

*Ports*: A Cross-

sectional Data Envelopment Analysis" *International Journal of Logistics:

Research and *

*Applications *9(1), pp. 19–31.

[13]

Fare R.S., Grosskopf S., Lovel C.A.K.. (1994). Production Frontiers.

Cambridge University

Press, Cambridge.

[14]

Haralambides, H.E., A. Verbeke, E. Musso and M. Benacchio, 2001. "Port

Financing and

Pricing in the European Union: Theory, Politics and Reality. *International

Journal of Maritime *

*Economics *3, pp. 368–386.

[15]

Lee Chee Xui, 2005. "Malaysian Container Port Performance Measurement" *M.Sc

Thesis*,

Malaysia University of Science and Technology.

[16]

Mahadevan, R., 2002. "A DEA Approach to Understanding the Produc¬tivity

Growth of

Malaysia's Manufacturing Industries" *Asia Pacific Journal of Management *19,

pp. 587–600.

[17]

Park, R.K. and P. De, 2004. "An Alternative Approach to Efficiency

Measurement of Seaports"

*Maritime Economics & Logistics *6, pp. 53–69.

[18]

Rios, L.R. and A.C.G Maçada, 2006. "Analysing the Relative Efficiency of

Container

Terminals of Mercosur *using* DEA" *Maritime Economics & Logistics *8(4),

pp. 331–346.

[19]

Tongzon, J, 2005. "Port Privatization, Efficiency and Competitiveness: Some

Empirical

Evidence from Container *Ports* (Terminals)" *Transportation Research Part A

– Policy and *

*Practice *39, pp. 405–424.

[20]

Trujillo, L. and B. Tovar, 2007. "The European Port Industry: An Analysis of

its Efficiency"

*Maritime Economics and Logistics *9(2), pp. 148–171.

[21]

Wang T., K. Cullinane, 2006. "The Efficiency of European Container Terminals

and

Implications for Supply Chain Management" *Maritime Economics & Logistics *8,

pp. 82–99.

------------------------------

*Page 16*

Efficiency of Middle Eastern and East African Seaports: Application of

DEA *Using* Window Analysis

613

[22]

William W. Cooper, Seiford L. M, Kaoru Tone (2002). DATA ENVELOPMENT

ANALYSIS.

Kluwer Academic Publishers. New York, Boston, Dordrecht, London, Moscow.

Over the past few decades, seaport industry in many countries of the world

has witnessed remarkable development. This is obvious, particularly in the

East African countries, such as Sudan, Eritrea, Djibouti, Kenya, and

Tanzania, and the Middle Eastern countries, particularly Saudi Arabia,

Yemen, Oman, the United Arab Emirates, and Iran. These countries possess

seaports which are strategically located in the international maritime trade

route between the East and the West (Figure 1) and are considered as middle

distance seaports. Goods carried from Europe and Far East/Australia and vice

versa can be exchanged and transhipped to all countries in the Middle East,

Red Sea, and East Africa. Since the olden days, these seaports have provided

services for the regional coasters and as time wentby, they have developed

to be among the important maritime international trade centres in the

region. The geographically strategic location of some of these seaports,

have also encouraged modern container vessels to make short duration calls

upon them (e.g. shipping lines operating along Asia/Europe route,

Asia/Mediterranean route and Asia/US East Coast route). These seaports and

their characteristics are displayed in Table 1.

http://www.eurojournals.com/ejsr_23_4_10.pdf

----[This List to be used for Eritrea Related News Only]----

© Copyright DEHAI-Eritrea OnLine, 1993-2009

All rights reserved